Repetitive transcranial magnetic stimulation (rTMS), when delivered to the dorsolateral prefrontal cortex (DLPFC) is a clinically effective treatment fo major depressive disorder (MDD). Network models of MDD are increasingly gaining acceptance, and functional connectivity MRI (fcMRI) has revealed topographically specific aberrations in functional network architecture in MDD. And yet, despite hints that the therapeutic effects of rTMS are actuated through distributed impacts upon cortical and subcortical limbic centers, the network effects of rTMS remain mysterious. This proposal seeks to investigate the way rTMS to DLPFC modulates network functional connectivity between the site of stimulation and a critical limbic region, the subgenual cingulate (sgACC), and between the sgACC and other limbic regions. My strategy is to use high frequency rTMS to stimulate the left DLPFC (recapitulating the therapeutic methodology), but to specifically target DLPFC regions that are functionally correlated and anti-correlated with the sgACC. This will be accomplished in a group of patients with MDD, and in a group of carefully matched controls. These subjects will be scanned before and after rTMS is delivered, as a way of gauging its effects. The feasibility of this design was recently demonstrated by our group (Eldaief et al. PNAS 2011). Changes induced by the stimulation will be charted with a novel combined MRI-PET (Positron Emission Tomography) scanner, which is capable of simultaneously recording fcMRI BOLD (Blood oxygenation level-dependent) and 18Flurodeoxyglucose (FDG) PET data. This will permit cortico-limbic networks to be characterized dynamically through (1) examination of the differential effects of stimulating two networks in the DLPFC, and (2) by tracking the dynamic interplay between rTMS induced changes in local glucose metabolism at DLPFC and sgACC on the one hand, and changes in distributed connectivity between these regions on the other. In an exploratory aim, MDD patients will return three months later (after they have undergone a non-specific treatment intervention with their psychiatric provider) for the identical rTMS/fcMRI/FDG-PET procedures. This aim will establish, as proof of principle, that treatment of MDD is associated with changes in cortico-limbic functional network architecture, and in cortico-limbic dynamics. Eventually, it is hoped that this work will lead to the emergence of aberrant cortical dynamics as a biomarker for MDD. In addition, this work might pollinate future studies which use aberrant cortical dynamics as a novel therapeutic target upon which neuromodulatory interventions might intervene.